LDAI-ISPS: LncRNA-Disease Associations Inference Based on Integrated Space Projection Scores

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Long non-coding RNAs (lncRNAs) play roles in cellular development and disease but are not translated into proteins.
• Current computational methods for inferring lncRNA-disease associations lack predictive accuracy due to issues with similarity networks and data limitations.

Purpose of the Study:

• To develop an improved computational method for inferring lncRNA-disease associations.
• To enhance the understanding of disease pathogenesis by accurately identifying related lncRNAs.

Main Methods:

• Proposed the lncRNA-disease associations inference based on integrated space projection scores (LDAI-ISPS) method.
• Constructed weighted networks by integrating disease semantic similarities, lncRNA functional similarities, and known lncRNA-disease associations.
• Utilized vector projections to obtain space projection scores for unbiased prediction.

Main Results:

• LDAI-ISPS achieved high accuracy in cross-validation: AUC of 0.9154 for inferring diseases, 0.8865 for inferring new lncRNAs, and 0.7518 for inferring isolated diseases.
• Demonstrated superior performance compared to existing methods.

Conclusions:

• LDAI-ISPS offers a more accurate and unbiased approach for predicting lncRNA-disease associations.
• The method serves as a valuable tool to complement traditional biological experiments in identifying potential lncRNA-disease relationships and isolated diseases.